Slightly
Relaxation of strained SiGe (%)
Slightly
Relaxation of strained SiGe (%)
Slightly
Relaxation of strained SiGe (%)
Slightly
Relaxation of strained SiGe (%)
Slightly
Relaxation of strained SiGe (%)
Slightly
Figure 3.4.3 Bow height change of four Si wafers with following Si1-xGex
deposition (x=0.2, 0.27, 0.33 and 0.44), implant of Arsenic, spike rapid anneal (RTA), and laser MSA annealing.
Figure 3.4.4 Correlation between the relaxation of strained SiGe and wafer bow height after laser MSA annealing.
80
wafer would cause about 35% relaxation of strained SiGe and induce large wafer bending during MSA.
3.4.3 Post-SiGe soak anneal effect on relaxation of strained SiGe
Fig.3.4.5 reveals the post-SiGe soak annealing effect on relaxation of strained Si1-xGex
(x~0.35) and its interaction with MSA causing dramatic wafer bow change. Post-SiGe implant causes damage at the upper portion of strained SiGe layer, instead post-SiGe soak anneal makes strained-SiGe relaxed by forming misfit dislocations at interface of SiGe/Si substrate. The four SiGe samples with the subsequent soak annealing 60sec, 600sec, 900sec, and 1200sec at 800°C caused 12%, 35%, 63%, and 74% relaxation of strained SiGe, respectively, then were directly followed by MSA to study the phenomena of large wafer bending and defect formation. The strained SiGe wafers with soak annealing 800°C of 60sec resulted in 12% relaxation of strained SiGe. Upon MSA, bow height was measured -15um in compressive state. No defect formation was observed in the underlying Si substrate as shown in the TEM picture of Fig. 3.4.6 (a). The strained SiGe wafer with soak annealing 800°C of 900sec resulted in 63% relaxation with SiGe surface migration. Upon MSA, bow height turned to high tensile state, up to 415um. The TEM image in Fig.3.4.6 (b) obviously shows the defect formation not only in SiGe film but also in underlying Si substrate. Finally, a longer soak annealing time caused an almost 74% relaxation of strained-SiGe, but the post MSA
81
caused bow height to less tensile state with 365um. Hence, the relaxation at upper portion of strained SiGe layer by implant or relaxation from interface of SiGe/Si by post-soak anneal is not the key factor to cause wafer bow change during MSA, the relaxation degree of strained SiGe is the key to cause significant wafer bending during MSA. Therefore, thermal budget larger than 800°C/60sec to prevent relaxation over 10% is necessary to be carefully integrated in strained SiGe, soak annealing and following MSA processes.
0 20 40 60 80 100
Relaxation of strain SiGe (%)
(600sec)
Relaxation of strain SiGe (%)
(600sec)
Figure 3.4.5 Upon MSA, correlation of bow height change and post-SiGe various soak thermal annealing causing relaxation of strained-SiGe. Medium relaxation of strained-SiGe by post-SiGe soak annealing also caused defect formation in the underlying Si substrate.
82
Figure 3.4.6 upon MSA, different degree of strained-SiGe relaxation caused by post soak anneal (a) 12% relaxation of strained SiGe did not show defects in Si (b) 36% relaxation of strained SiGe showed defects in the underlying Si
Figure 3.4.7 The schematic plot of in-situ Ge content, soak anneal and implant effect on defect formation in the underlying Si during MSA
83
The schematic plot of in-situ Ge content, soak anneal and implant effect on defect formation in the underlying Si during MSA as shown in Figure 3.4.7. All of the three methods including in-situ Ge fraction, post-SiGe implant and thermal anneal caused relaxation of strained SiGe and defect formation in the underlying Si substrate during MSA. The key message we need to know is the relaxation degree of strained SiGe to cause defect formation, instead of what methods used to cause relaxation. Therefore, suppressing the total relaxation level of strained SiGe below 10% is important to avoid defect formation in Si for 28nm PMOSFETs volume production.
3.5 Summary
We have concluded that the degree of strained-SiGe relaxation importantly caused wafer bending as well as the defect formation in the underlying Si substrate when MSA was applied to the relaxed strained-SiGe wafers. In either cases of post-SiGe implant or post-SiGe soak annealing or in-situ Ge at.% resulting in relaxation of strained SiGe, low relaxation less than 10% did not cause large SiGe wafer bending during MSA. Upon MSA medium relaxation level indeed caused significant SiGe wafer bow height change from compressive to high tensile state as well as defect formation in the underlying Si substrate. The key message we need to know is the relaxation degree of strained SiGe to cause defect formation, instead of
84
what methods used to cause relaxation. Therefore, suppressing the total relaxation level of strained SiGe below 10% is important to avoid defect formation in Si for 28nm PMOSFETs volume production.
85
References
[1]. Luo, Z.; Chong, Y.F.; Kim, J.; Rovedo, N.; Greene, B.; Panda, S., “Design of high performance PFETs with strained Si channel and laser anneal”,International Electron Devices Meeting, p. 489-492 (2005)
[2]. Cáit Ní Chléirigh, XiaoRu Wang, Gana Rimple, Yun Wang, N. David Theodore, Michael Canonico, and Judy L. Hoyt,” Super critical thickness SiGe-channel heterostructure p-type metal-oxide-semiconductor field-effect transistors using laser spike annealing”, Journal of Applied Physics volume 103, issue 104501 (2008)
[3]. Ghani, T.; Armstrong, M.; Auth, C.; Bost, M.; Charvat, P.; Glass, G.; Hoffmann, T,
“Strain Engineering for Enhanced P-channel Field Effect Transistor Performance”
International Electron Devices Meeting, p. 978-980 (2003)
[4]. Scott E. Thompson；Mark Armstrong；Chis Auth；Mohsen Alavi；Mark Buehler；Robert Chau, et al., “ A 90-nm logic technology featuring strained-silicon “ IEEE Transaction on Electron Devices, Vol. 51, p1790-1797 (2004)
[4]. Kenji Shibata, Yamichi Ohmura, Tomoyasu Inoue, Koichi Kato, Yasuhiro Horiike and Masahiro Kashiwagi, “Growth of Large Grained Silicon on Insulator by Electron Beam Annealing and Performance of MOS Devices”, Japanese Journal of Applied Physics. 21 pp. 1111-1116 (1982)
[5]. Lee, M.L. ; Yu, B. ; Lee, K.L. ; Dohrman, C.L. ; Isaacson, D. ; Langdo, T.A. ;
86
Antoniadis, D.A. “Dislocation engineering in strained mos materials”, International Electron Devices Meeting, p. 519 (2005)
[6]. Grace Huiqi Wang and Eng-Huat Toh, Xincai Wang, Sudhiranjan Tripathy, et al.,
“Strained SiGeSn formed by Sn implant into SiGe and pulsed laser annealing” Applied Physics Letters 91, 202105 (2007)
[7]. M. Bargallo Gonzalez, E. Simoen, E. Rosseel, P. Verheyen, L. Souriau, “Temperature profile during millisecond processing” ECS Transactions vol.13 no.1 (2008)
[8]. B. P. Fairand, A. H. Clauer, R. G. Jung, and B. A. Wilcox, “Quantitative assessment of laserinduced stress waves generated at confined surfaces”, Applied Physics Letters, 25, p431-433, (1974)
[9]. C.H. Chen; Nieh, C.F. ; Lin, D.W. ; Ku, K.C. ; Sheu, J.C. ; Yu, M.H. ; Wang, L.T. ; Lin, H.H. ; Chang, H. ; Lee, T.L. ; Goto, K. ; Diaz, C.H. ; Chen, S.C. ; Liang, M.S. “Channel Stress Modulation and Pattern Loading Effect Minimization of Milli-Second Super Anneal for Sub-65nm High Performance SiGe CMOS” P174 – 175, VLSI Technology, (2006).
[10]. M. H. Yu, L. T. Wang, T. C. Huang, T. L. Lee, and H. C. Cheng, “The Strained-SiGe Relaxation Induced Underlying Si Defects Following the Millisecond Annealing for the 32 nm PMOSFETs” Journal of The Electrochemical Society 159 (2012) H243
87